International Journal for Research in Applied Science & Engineering Technology (IJRASET)

ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor:6.887

Volume 5 Issue VIII, August 2017- Available at www.ijraset.com

A Proposed Technique for Privacy Preservation by

Anonymization Method Accomplishing Concept of
K-Means Clustering and DES
Priyanka Pachauri1, Unmukh Datta2
1
Dept. of CSE/IT, MPCT College Gwalior, India
2
Prof. Dept. of CSE/IT, MPCT College , Gwalior, India

Abstract: Privacy-Preserving Data Mining also an essential branch of the data mining and an exciting topic in privacy
preservation has gain particular attention in current years. Data mining has been substantially studied and useful into numerous
fields which include the Internet of Things and the business growth. However, data mining tactics additionally take vicinity
critical demanding situations due to enlarged sensitive statistics disclosure and the violation of privacy. This discussion describes
the privacy concern that occurs due to data mining, particularly for the national security applications. We discuss PPDM by
Anonymization Method in which we use K-means clustering in order to divide the given data and DES algorithm for encryption
of data in order to prevent sensitive data from attacker.
Keywords: privacy preservation, PPDM, K-meanS clustering, anonymization ,DES etc.

I. INTRODUCTION
Privacy is receiving extra attention partially as of counter-terrorism and the national security. At present we have heard so much
approximately countrywide protection in media. This is mainly because people are now realizing that to handle terrorism, the
government may need to collect information about individuals. There been much interest in the recent on applying the sa mining
used to detect patterns which are unusual , terrorist activities and the fraudulent behavior. While all applications of data mining can
give profit to the humans and save lives, there also negative side to this type of method , it could be a danger to the individuals
privacy . This is due to data mining tools are present on the Web or, and even naive individuals can use these tools to mine
information from stored data in various databases and files, and therefore violate the privateness of individuals. As we have
stressed in papers to take out efficient data mining and mine necessary information for counter terrorism and national security, we
gather all kinds of information about individuals.[1] However, this information could be a threat to individuals’ privacy and civil
liberties. This is causing major concern with different civil liberties unions. The aim is to carry out data mining and yet to the
maintain privacy. This topic is known as privacy-preserving data mining .
In this paper we use anonymization technique along with K-means clustering and DES algorithm.

A. Anonymization
To guard identity of the individual when release of sensitive information is done , holders of data often being encrypt or eliminate
explicit identifiers, like names and the unique security no: However, data which is unencrypted provides no assurance for
anonymity. To preserve privacy, model of k-anonymity has been proposed by the Sweeney [2] which achieve k anonymity by
means of generalization and the Suppression , K-anonymity, it is difficult for an imposter to decide the identity of the individuals in
collection of data set containing personal information. For ex, the age of person might be generalized to a variety like youth, middle
age and the adult with no specifying suitably, so as to decrease the threat of the identification. Suppression involves decrease the
exactness of the applications and it don’t liberate some information .By using this method it reduces the risk of detecting exact
information.
B. K- Means Clustering Algorithm
K-mean is the greatest popular apportioning procedure of clustering. It turns out to be leading proposed by methods for Macqueen in
1967. K-mean is an unsupervised, non-deterministic, numerical, iterative technique for grouping. In k-mean each cluster is
represented by using the suggest value of gadgets within the cluster.[3] Here we partition a hard and fast of n item into k cluster in
order that intercluster similarity is low and intracluster similarity is excessive. Similarity is measured in time period of imply value
of objects in a cluster. The set of rules consists of two separate levels.

©IJRASET (UGC Approved Journal): All Rights are Reserved 1594

 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor:6.887
Volume 5 Issue VIII, August 2017- Available at www.ijraset.com

1 st Phase pick k centroid randomly, where the value k is settled in advance.2 nd Phase each item in data set is identified with the
closest centroid. Euclidean distance is used to measure the distance among every data object and cluster centroid.
Procedure
1) Arbitrarily pick out k data item from D dataset as preliminary cluster centriod.
2) Repeat
3) Assign every data item to the cluster to which object is most comparative construct absolutely with respect to the suggest
estimation of the item in cluster;
4) Calculate the new suggest estimation of the data items for each cluster and refresh the mean esteem.
5) Until no trade.
C. DES(Data Encryption Standard)
Data Encryption Standard also called as (DES) algorithm has been all the rage secret key encryption algo and it is taken in use in
lots of commercial and the financial applications. Though it was introduced in the year 1976, it has established resistant to all the
forms of cryptanalysis. In addition, DES is a block cipher algorithm which means that it takes a fixed length of the message and
encrypts it (encrypts the block), and returns a string in the same size. DES is the first encryption algorithm recommended by NIST
(National Institute of Standards and Technology).

II. LITERATURE REVIEW

The main contributions of this paper are three folds: (i) the definition of the data collection and publication process, (ii) the privacy
framework model and (iii) personalized anonymization approach. The experimental analysis is presented at the end; it shows this
approach performs better over the distinct l-diversity measure, probabilistic l-diversity measure and k-anonymity with t-closeness
measure [4].
The [5] proposes a singular, extra flexible generalization scheme. The experimental results of their study indicate that their
approaches produce k – anonymization with less generalization compared to previous approaches. They conclude that a bottom-up
approach for k – anonymization is preferable for small number of quasiidentifying attributes.
The k anonymity based method is illustrated in [6] is used to search for optimal feature set partitioning and [7] for cluster analysis.
And [8] Proposes a data reconstruction approach to obtain k-anonymity safety in predictive data mining. In this method the probably
identifying attributes are first mapped the usage of aggregation for numeric data and swapping for nominal data. A genetic set of
rules technique is then implemented to the masked data to find a correct subset of it.
k-anonymity method is treated as the classical anonymization method and most of the studies are based on k-anonymity. The others
are based on its improved methods like l-diversity, t-closeness, km -anonymization, (α,k) anonymity, p-sensitive k-anonymity, (k,e)
anonymity, which are described in [9].
They provide a detailed survey of anonymization methods and also point out pitfalls in k anonymity. Previous works by Samarati
and Sweeney [10,11] shows that the removal of the personally identifying information from data is insufficient for the data security,
rather it is better to use k – anonymity method for publishing data. The quasi–identifier (QI), which is the combination of person
specific identifiers are considered here for the process of anonymization. One of the common methods to achieve k –anonymity is to
generalize identifiers (for example date of birth can be generalized to month of birth).
[12]There are various PPDM techniques such as anonymization, perturbation, randomization, condensation, cryptography. In this
paper we have reviewed anonymization technique of PPDM such as k anonymity using generalization and suppression, p sensitive k
anonymity, (α, k) anonymity, l diversity, m-invariance.
Sometimes the data should be publically published in its original form. Even though it is not encrypted and perturbed, some sort of
precaution should be implemented before releasing the data in terms of anonymization [13]. This is a kind of generalization of some
attributes which protects against identity disclosure. Anonymization can be obtained through techniques inclusive of generalization,
suppression, data removal, permutation, swapping and so forth [13]
III. PROPOSED METHOD
The proposed algorithm is a try to present a new method for complex encrypting and decrypting data based totally on parallel
programming in the sort of way that the new method can make use of multiple- core processor to acquire higher speed with better
degree of protection.

©IJRASET (UGC Approved Journal): All Rights are Reserved 1595

 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor:6.887
Volume 5 Issue VIII, August 2017- Available at www.ijraset.com

ALGORITHM
Partition of given dataset through the usage of K- means clustering
Summary  partition
Dim  choose dimension()
P  frequency set(partition , dim)
Sv  find median(P)
Lhs  {t belongs to partition : t.dim <= sv}
Rhs  {t belongs to partition: t.dim > sv}
Apply DES for encrypting the records.
Finally return union of lhs and rhs partition

start

Input dataset

Anonymization algorithm

K-Means clustering for partitioning

dataset

DES encryption on data

Final result

end

Fig. 1 Flow diagram Propose method

Procedure –
Step1 - consider the dataset for input.
Step2 - applies anonymization technique to that particular dataset.
Step3 - K-means clustering technique is used to partiton the data sets into clusters.
Step4 – DES encryption technique is used to suppress the data values.
Step5 – final result obtained by union of lhs and rhs values formed by anonymization technique.

IV. RESULT ANALYSIS

Table 1 describes the accuracy between Base method results and Propose method results.

No. of 100 200 300 400 500

records
Accuracy 84.00 83.50 84.00 84.00 83.80
in base
results
Accuracy 96.00 98.00 99.00 98.75 98.20
in
proposed
results

©IJRASET (UGC Approved Journal): All Rights are Reserved 1596

 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor:6.887
Volume 5 Issue VIII, August 2017- Available at www.ijraset.com

Fig.2 Above graph represents the analysis of accuracy in base and propose method results and concluded that propose method is
best for preserving privacy.

Table II describes the error rate between Base method results and Propose method results.

No. of 100 200 300 400 500

records
Accuracy 84.00 83.50 84.00 84.00 83.80
in base
results
Accuracy 96.00 98.00 99.00 98.75 98.20
in
proposed
results

Fig.3 Above graph represents the analysis of error rate in base and propose method results and concluded that propose method is
best for preserving privacy as the error rate in propose method is less than the error rate in base method.

V. CONCLUSION
Each privacy preserving technique has its own importance. Substantial efforts have been accomplished to address these needs.
Data encryption and anonymization are widely adopted ways regarding privacy breach. However, encryption is not suitable for data
that are processed and shared. Anonymizing huge data and dealing with anonymized data sets are nonetheless challenges for classic
anonymization processes. Privacy- preserving data mining is emerged for to 2 critical desires: data analysis with a purpose to deliver
better services and making sure the protection privileges of the data owners.

©IJRASET (UGC Approved Journal): All Rights are Reserved 1597

 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor:6.887
Volume 5 Issue VIII, August 2017- Available at www.ijraset.com

The results of our proposed work shows that by doing k-means clustering and encypting the data using DES method we can achieve
more preservation of privacy.

REFERENCES
[1] Bhavani Thuraisingham,” Privacy-Preserving Data Mining: Developments and Directions”, IDEA GROUP PUBLISHING, Journal of Database Management,
16(1), 75-87, Jan-March 2005 77.
[2] Pingshui WANG,” Survey on Privacy Preserving Data Mining”, International Journal of Digital Content Technology and its Applications, Volume 4, Number
9, December 2010.
[3] Jyoti Yadav, Monika Sharma,” A Review of K-mean Algorithm”, International Journal of Engineering Trends and Technology (IJETT) – Volume 4 Issue 7-
July 2013, ISSN: 2231-5381.
[4] M. Prakash G. Singarave “An approach for prevention of privacy breach and information leakage in sensitive data mining”2015 IEEE.
[5] Tiancheng Li, Ninghui Li, “Towards Optimal k-anonymization”, Data & Knowledge Engineering, 2008 Elsevier.
[6] Nissim Matatov, Lior Rokach, Oded Maimon, “Privacy-preserving data mining: A feature set partitioning approach”, Information Sciences 180 (2010) 2696–
2720.
[7] Benjamin C. M. Fung, Ke Wang, Lingyu Wang, Patrick C.K. Hung, “Privacy-preserving data publishing for cluster analysis” , Data & Knowledge Engineering
68 (2009) 552–575.
[8] Dan Zhu, Xiao-Bai Li, Shuning Wu, “Identity disclosure protection: A data reconstruction approach for privacypreserving data mining”, Decision Support
Systems 48 (2009) 133–140.
[9] Yan Zhao, Ming Du, Jiajin Le, Yongcheng Luo, “A Survey on Privacy Preserving Approaches in Data Publishing”, First International Workshop on Database
Technology and Applications, 2009.
[10] Samarati P, “Protecting respondent’s privacy in Microdata release”, IEEE Transactions on Knowledge and Data Engineering, 13:1010–1027.
[11] Sweeney L, “k-anonymity: A model for protecting Privacy”, International Journal on Uncertainty, Fuzziness and Knowledge-based Systems, 10(5):557–570.
[12] Kiran Israni, Shalu Chopra,” Survey on Anonymization Technique for Privacy Preserving Data Mining (PPDM)”, International Journal of Innovative Research
in Computer and Communication Engineering, Vol. 4, Issue 11, November 2016, ISSN(Online): 2320-9801.
[13] Asmaa H.Rashid and Prof.dr. Abd-Fatth Hegazy, “Protect Privacy of Medical Informatics using K-Anonymization Model”, IEEE Explore.

©IJRASET (UGC Approved Journal): All Rights are Reserved 1598